Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In with OpenID
Advertise on LowEndTalk.com

In this Discussion

What you pay may not be what you get in cloud VPS?

What you pay may not be what you get in cloud VPS?

96mb96mb Member
edited October 2012 in General

A similar topic has been posted on WHT, but I think it might be a good idea to ask a different question on LET since it represents more the voice of the low end providers, so here it goes:

I used to be a firm believer of what-you-pay-is-what-you-get (in the cloud VPS hosting market, since everyone's set up is pretty much the same in terms of cost), however, after reviewing some of the cloud VPS providers that I have came across in the recent days, I think I am no longer very certain about this.

For example, with the three most recent VPS reviews that I have done:

AtomicVPS one node, which costs 40 per month:

http://www.96mb.com/96mb-low-end-vps-review-part-57atomic-vps/

VPS.NET one node, which costs half of that:

http://www.96mb.com/96mb-low-end-vps-review-part-55-vps-net/

And VMStorm VPS, which costs a little more than a third of what VPS.NET costs per month:

http://www.96mb.com/96mb-low-end-vps-review-part-56vm-storm-vps/

As you can see, the performance of VMStorm VPS is clearly better than VPS.NET one node which in turn is clearly better than AtomicVPS.

I am not trying to bash any provider here and granted the higher plans of VPS.NET and AtomicVPS might show a lot better performance (and AtomicVPS has indicated they have 100% uptime in August: https://twitter.com/atomicvps/status/241985680731959296).

So here is the question (which is different from what I asked on WHT):

Without divulging your business secrets, what do you think are the differences between you, as a low end VPS provider and the providers who charge many times more than you do and do you think why would people still willing to pay many times more to get a VPS from the high end providers when they could get comparable, if not better performance from one of the LEB providers?

96MB: My Low End VPS Review Blog, 96MB Forum: Official forum for 96mb.com, LowEndPress: Low End VPS community news/tips site on 64MB of RAM!

Comments

  • Well "Cloud" can mean lots of think. Like amazon ec2 you can deploy so much server in minutes, but there nodes fail at times. So need to know what your doing. "Cloud can also mean uptime like it moves vps in real time when node is about to go down. There isn't true meaning to cloud to me.

  • PatrickPatrick Member
    edited October 2012

    @96mb said: Without divulging your business secrets, what do you think are the differences between you, as a low end VPS provider and the providers who charge many times more than you do and do you think why would people still willing to pay many times more to get a VPS from the high end providers when they could get comparable, if not better performance from one of the LEB providers?

    People prefer to pay more to known brands who can deliver, they mostly pay for better service in terms of support and know your in good hands then better performance or cheaper prices.

  • I'm not a VPS host, but in my experience, the reason people buy more expensive services is exactly because of the "You get what you pay for" mantra. They believe that expensive services must be better than cheap services - or even claim that it's "impossible" to offer services for "those prices" (refering to LEB pricing).

    Appreciate my posts/software/guides? Donate (PayPal/Flattr/Bitcoin): http://cryto.net/~joepie91/donate.html | irc.freenode.net #lowendbox

    Thanked by 196mb
  • 24khost24khost Member
    edited October 2012

    Most "cloud" providers over promise and under deliver, Like with us and most of the lowend cloud providers we believe in under promising and over delivering.

  • It's good this question was raised. There is a really big confusion out there on what a cloud really is, also because there is no real official definition of cloud. In my opinion a real cloud should be made of the following minimal components:

    -1 control panel node -2 hypervisors nodes -SAN Storage or 1 storage node

    I have been recently digging this cloud thing and what I found out is that many providers dont really deploy cloud servers out of a real cloud infrastrucutre, which means using the minimal setup above.

    Many cloud servers offered on the markets are provided using one single hypervisor node or also local storage. This especially applies to smaller providers. What really seems to make the difference is the software used to run the cloud. Maybe because in the end that's what client sees, not the infrastructure. Software like CloudStack and OpenStack can run perfectly on a single server, but it that a cloud? I would say no.

    Personally I dont find strange that a cloud VM costs on average 2 times a normal VPS considering that such a structure has to be redundant and so has to contain double hardware. I think clients should know if their cloud VM runs really on a cloud or not.

    VirtualSRV - OpenVZ Unmanaged VPS
  • @dragonballz2k said: So need to know what your doing. "Cloud can also mean uptime like it moves vps in real time when node is about to go down. There isn't true meaning to cloud to me.

    That is actually something puzzled me as well, if the cloud providers are HA providers, should automatic fail-over be sort of built-in features for them?

    96MB: My Low End VPS Review Blog, 96MB Forum: Official forum for 96mb.com, LowEndPress: Low End VPS community news/tips site on 64MB of RAM!
  • @96mb Yes auto fail-over is a must.

  • @96mb cloud has basically two meaning. Amazon ec2 or fail over but then to auto fail over you need a good automatic system.

  • @dragonballz2k correct and this is where like the system we use vpsgrid comes in handy, as it takes care of all the auto-failover and data-replication.

  • @24khost it can mess up easy if the company didn't do it right. Like what happens when all vps nodes are full. Amazon ec2 is not safest if not done right either like hardware could fail. It not as safe running one amazon ec2 instance.

  • Correct and that is why we use vps-grid designed by virtualcomplete. We don't have those problems.

  • edited October 2012

    So?

    Obviously different companies are going to charge differently. Personally, I'd rather charge more and manage a smaller fleet of clients than charge less and have a ton of clients to support.

    @96mb said: Without divulging your business secrets, what do you think are the differences between you, as a low end VPS provider and the providers who charge many times more than you do and do you think why would people still willing to pay many times more to get a VPS from the high end providers when they could get comparable, if not better performance from one of the LEB providers?

    I think the only main difference between a 'low end provider' and a 'high end provider' is who the provider is trying to target. Some really have no interest in the 'everything for nothing' crowd, even if it can be a sustainable business model. Don't get me wrong, I love LEB and everything but this isn't really the market I'd want to work hard in succeeding in.

    Furthermore, as I said above, I'd rather manage 100 clients paying $40/mo as opposed to 800 clients paying $5/mo. Especially because I think there would be less abuse of services with the $40/mo crowd.

    Just my $0.02

    vpsBoard.com - Now with over 450 members! A friendly community with active discussion. Come join us!

    IRC.FREENODE.NET #vpsBoard - Drop by and say, 'Hello'.
  • Responding to the "what is cloud" sub-thread: US NIST gives a definition here that has received pretty broad support: http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf.

    Obviously there's a ton of companies claiming their stuff is "cloud" when it isn't...but this shouldn't surprise anyone who has been in the IT industry for more than 5 nanoseconds as this happens regularly (just look at those claiming SDN right now...).

    "Go cheap on rarely used things"

  • MaouniqueMaounique Member
    edited October 2012

    I think it should have failover, separate storage of some kind, one panel and more than 2, I would say 4-5 hypervisor nodes. A scheme with only 2 would be a HA scheme of sorts as failover is a must in my view for a cloud, actually, it would be the main selling point for my case. As ppl said before, double the price would be logical since it should have supraunitary capacity, depending on the number of hypervisors and the scheme employed as well as other overheads,such as external storage and the quite expensive ways to connect to it in order to deliver a really good performance. I think the bigger the cloud and the more complex the HA scheme the more complicated the setup will be but will have a lower overhead as implementing HA will mean much less than double the capacity. I also think that setting up a really good cloud is a very complicated task, as I foresee many problems that will lead to downtime due to tweaks and mistakes because not many ppl have the first hand experience with a big and loaded cloud yet. eventually, the cloud structure of today will expand like the internet expanded from a few connected nets and we will be able to move our instances to float around, carriers will offer not only dataflows but also storage space, at least temporary. We are at most 10 years away from this, probably 5, we will be able to move our LEBs around the world with minimal hustle. It will be a monumental task, but a set of common APIs will have to be provided by everyone and lock-ins will probably be frowned upon everyone. The dreaded downtime will happen only to uprepared ppl, much like today ppl lose data because of a lack of backup, in the future there will be downtime because of a lack of failover or very bad failover provider that will be overselling slots betting dowtime wont happen to the bigboys or to a big part of the internet. M

    top - 22:32:38 up 906 days, 2:58, 1 user, load average: 3.94, 5.13, 8.38

Sign In or Register to comment.